$$\begin{gathered} {\mathrm{Li}}_2{\left(6{\mathrm{e}}^{-}\right) : \left(\mathrm{\sigma }1\mathrm{s}\right)}^2{\left({\mathrm{\sigma }}^{*}1\mathrm{s}\right)}^2{\left(\mathrm{\sigma }2\mathrm{s}\right)}^2 \\ {\mathrm{Li}}_2^{-} {\left(7{\mathrm{e}}^{-}\right) : \left(\mathrm{\sigma }1\mathrm{s}\right)}^2{\left({\mathrm{\sigma }}^{*}1\mathrm{s}\right)}^2{\left(\mathrm{\sigma }2\mathrm{s}\right)}^2{\left(\mathrm{\pi }2{\mathrm{p}}_{\mathrm{x}}\right)}^1 {\mathrm{Li}}_2^{+} {(5\mathrm{e}}^{} \end{gathered}$$

Li²⁺​ is more stable than Li²⁻​ because Li²⁻​ has more numbers of antibonding electrons. More the bonding electrons, more is the stability.

Hence, Li₂^- < Li₂⁺ < Li₂